BACKGROUND OF THE INVENTION
1. Field of the invention
The present invention relates to an exposure
apparatus, and the fabrication method of a
semiconductor device using the same. More
particularly, the present invention relates to the
alignment of a mask with a wafer in a projection
exposure apparatus having a projection optical system
mounted thereon, which is a catadioptric type
including a reflecting member.
2. Related Background art
Conventionally, a projection exposure apparatus
has been used for fabricating a semiconductor device,
a flat panel display device (e.g., a liquid crystal
display device), and so on in a lithography process.
With a great increase made in the progress speed of a
semiconductor technology in recent years, a progress
in a microfabrication technology has been remarkable.
For example, in a semiconductor device having a
memory enabling writing/reading to be performed as
occasion demands, i.e., Dynamic Random Access Memory
(DRAM), currently a 16 M bit DRAM is a mainstream,
but developments are under way to increase the degree
of integration to the range of 256 M bit DRAM.
With such integration of the semiconductor
device, an exposure wavelength has been made shorter
than a currently mainstream i line (365 nm). In
other words, a KrF excimer laser (oscillation
wavelength 248 nm) has already been put to practical
use. In addition, developments are under way to
realize the practical use of an ArF excimer laser
(oscillation wavelength 193 nm).
The shorter exposure wavelength causes a
reduction in the transmittance of an optical material,
and a great limitation is placed on the kinds of
optical materials used for, for example a projection
optical system. Consequently, in the conventionally
mainstream exposure apparatus having the projection
optical system of a catadioptric type, inconvenience
inevitably occurs, i.e., the axial (longitudinal)
chromatic aberration of the projection optical system
becomes relatively large.
On the other hand, in the case of the exposure
apparatus having the projection optical system of a
catadioptric type, it is possible to suppress the
axial chromatic aberration of the projection optical
system even for the optical materials of limited
kinds. However, in the projection optical system of
the catadioptric type, because its optical axis is
bent by the reflecting member without being linearly
extended, inconvenience is inherent, i.e., the
fluctuation of a mask pattern image (positional shift
or rotational shift of an image) caused by the
position fluctuation of the reflecting member occurs
far more easily than that in the projection optical
system of the refractive type. The projection
optical system of the refractive type means a
projection optical system including no reflecting
members but a refraction system like a lens. The
projection optical system of the catadioptric type
means a projection optical system including at least
one reflecting member and a refraction system like a
lens.
Each of Japanese Patent Laid-Open No. 63-41023
and Japanese Patent Laid-Open No. 7-22350 disclose
the system of alignment between a mask and a wafer
(photosensitive substrate) in an exposure apparatus
having the projection optical system of a
catadioptric type mounted thereon. In the exposure
apparatus disclosed in each of these publications, a
beam splitter like a half prism is added in an
optical path between the mask and the wafer, and the
positions of the mask and the wafer relative to each
other are measured by making an alignment light
incident on the half prism.
Japanese Patent Laid-Open No. 5-21314 discloses
a method of measuring the positions of a mask and a
wafer relative to each other before exposure by using
a reference mark formed on a wafer stage. This
method is applied to the exposure apparatus of a
scanning type (step-and-scan system) for transferring
a mask pattern to each exposure area of the wafer
while moving the mask and the wafer relative to each
other with respect to the projection optical system
of a refractive type.
In the exposure apparatus disclosed in each of
Japanese Patent Laid-Open No. 63-41023 and Japanese
Patent Laid-Open No. 7-22350, the half prism is added
in the optical path between the mask and the wafer.
In other words, in the optical path between the mask
and the wafer, the half prism, not an essential
component of the projection optical system, is
disposed in surplus. Consequently, the optical
performance of the projection optical system is
adversely affected easily by the added half prism.
In addition, since the optical separation surface
(i.e., dividing surface of wave front) of the half
prism cannot be corrected after its fabrication,
excessively high accuracy is required for the
fabrication of the half prism to prevent the optical
performance of the projection optical system from
being adversely affected.
On the other hand, in the case of the exposure
apparatus described in Japanese Patent Laid-Open No.
5-21314, the positions of the mask and the wafer
relative to each other are measured before exposure
by using the reference mark formed on the wafer stage.
In other words, the positions of the mask and the
wafer relative to each other cannot be measured
during exposure as occasion demands. Consequently,
if the alignment system described in this publication
is applied to the exposure apparatus having the
projection optical system of the catadioptric type,
even when the position of the reflecting member in
the projection optical system changed during exposure
and causes a mask pattern image to be shifted in
position or rotation from a reference image-forming
position, alignment between the mask and the wafer
cannot be carried out for the projection optical
system during exposure according to the positional or
rotational shift of the mask pattern image as
occasion demands.
DISCLOSURE OF THE INVENTION
The present invention is made with the
foregoing problems in consideration. The object of
the present invention is to provide an exposure
apparatus capable of performing alignment between a
mask and a wafer even during exposure as occasion
demands, without providing any surplus optical
members in an optical path between the mask and the
wafer, and according to the positional or rational
shift of a mask pattern image caused by the position
fluctuation of a reflecting member disposed in the
projection optical system of a catadioptric type.
Another object of the present invention is to provide
a fabrication method of a semiconductor device using
the exposure apparatus.
In order to solve the foregoing problems, in
accordance with a first aspect of the present
invention, an exposure apparatus is provided,
comprising: an illumination optical system for
illuminating a mask having a specified pattern formed
thereon; a projection optical system for projecting
an image of the pattern on the mask to a
photosensitive substrate through at least one
reflecting member; a detection system for detecting
an amount of fluctuation from the reference position
of the at least one reflecting member; an arithmetic
system for computing an amount of correction
regarding at least one of the mask and the
photosensitive substrate, to substantially align the
pattern image formed in the state of being moved from
a reference image-forming position with the
photosensitive substrate, based on the amount of
fluctuation detected by the detection system; and a
driving system for moving at least one of the mask
and the photosensitive substrate based on the amount
of correction computed by the arithmetic system.
According to the preferred embodiment of the
first aspect of the present invention, the arithmetic
system computes the amount of positional shift and
the amount of rotational shift of the pattern image
from the reference image-forming position based on
the amount of fluctuation detected by the detection
system, and then computes the correction amount of
the mask necessary for substantially correcting the
amount of positional shift and the amount of
rotational shift. The driving system moves only the
mask based on the amount of correction computed by
the arithmetic system. Alternatively, the arithmetic
system computes the amount of positional shift and
the amount of rotational shift of the pattern image
from the reference image-forming position based on
the amount of fluctuation detected by the detection
system, and then computes the correction amount of
the photosensitive substrate necessary for
substantially aligning the photosensitive substrate
with the pattern image formed in a state where at
least one of the amount of positional shift and the
amount of rotational shift occurs. The driving
system moves only the photosensitive substrate based
on the amount of correction computed by the
arithmetic system. Otherwise, the arithmetic system
computes the amount of positional shift and the
amount of rotational shift of the pattern image from
the reference image-forming position based on the
amount of fluctuation detected by the detection
system, then computes the correction amount of the
mask necessary for substantially correcting one of
the amount of positional shift and the amount of
rotational shift, and still yet computes the
correction amount of the photosensitive substrate
necessary for substantially aligning the
photosensitive substrate with the pattern image in a
state where the other of the amount of positional
shift and the amount of rotational shift occurs. The
driving system moves the mask and the photosensitive
substrate based on the correction amounts of the mask
and the photosensitive substrate computed by the
arithmetic system.
In accordance with a second aspect of the
present invention, an exposure method is provided to
project and expose a pattern image provided on a mask
onto a workpiece through a projection optical system
having at least one reflecting member.
In this case, the exposure method comprises: a
detection step of detecting a fluctuation amount of
the at least one reflecting member from a reference
position; an arithmetic step of computing an amount
of correction regarding at least one of the mask and
a photosensitive substrate, necessary for
substantially aligning the photosensitive substrate
with the pattern image formed in a state of being
moved from a reference image-forming position, based
on the amount of fluctuation detected in the
detection step; an alignment step of performing
alignment between the mask and the photosensitive
substrate with respect to the projection optical
system by moving at least one of the mask and the
photosensitive substrate based on the amount of
correction computed in the arithmetic step; and an
exposure step of illuminating the mask by an
illumination optical system in a state where the mask
and the photosensitive substrate are aligned with
each other with respect to the projection optical
system in the alignment step, and exposing a pattern
of the mask on the photosensitive substrate through
the projection optical system.
According to the preferred embodiment of the
second aspect of the present invention, the detection
step, the arithmetic step and the alignment step
should preferably be performed before the exposure
step. The detection step, the arithmetic step and
the alignment step should preferably be performed
during the exposure step as occasion demands. The
exposure method should further comprise, preferably,
a determination step of determining whether the
fluctuation amount of the at least one reflecting
member from the reference position is permissible or
not and, if non-permissibility thereof is determined,
then it is preferable that the arithmetic step and
the alignment step are performed. The exposure
apparatus should further comprise, preferably, a
determination step of determining whether the
fluctuation amount of the at least one reflecting
member from the reference position is permissible or
not and, if non-permissibility thereof is determined,
then it is preferable that information regarding the
amount of fluctuation is displayed. In addition, in
the arithmetic step, an amount of correction should
preferably be computed for at least one of the mask
and the photosensitive substrate, which is necessary
for substantial alignment between the mask and the
photosensitive substrate with respect to the
projection optical system. The computation is
performed based on the movement amount of the pattern
image from the reference image-forming position,
which is obtained based on the amount of fluctuation
detected in the detection step, and based on the
position information of a mask stage for holding the
mask and movable with respect to the projection
optical system, and based on the position information
of a substrate stage for holding the photosensitive
substrate and movable with respect to the projection
optical system.
Furthermore, in the first and second aspects of
the present invention, the at least one reflecting
member should preferably be a reflecting mirror
having no power (refracting power).
BRIEF DESCRIPTION OF THE DRAWINGS
Fig. 1 is a schematic view showing a
constitution of an exposure apparatus according to a
first embodiment of the present invention.
Fig. 2 is a view showing a state of rotational
fluctuation of a first reflecting mirror M1 shown in
Fig. 1 from a reference position clockwise around a Z
axis.
Fig. 3 is a view showing an entire movement of
an exposure field EF and movements of image points P1
and P2 when rotational fluctuation occurs in the
first reflecting mirror M1.
Fig. 4 is a view showing a state of rotational
fluctuation of a second reflecting mirror M2 shown in
Fig. 1 from the reference position clockwise around a
Y axis.
Fig. 5 is a view showing an entire movement of
the exposure field EF and movements of the image
points P1 and P2 when rotational fluctuation occurs
in the second reflecting mirror M2.
Figs. 6A to 6C are schematic perspective views,
each of them showing a specific constitution of a
detection system for detecting a rotational
fluctuation angle of a reflecting mirror.
Fig. 7 is a flowchart illustrating processes
for forming a specified circuit pattern on a wafer by
using the exposure apparatus of the present invention.
Fig. 8 is a schematic view showing a
constitution of an exposure apparatus according to a
second embodiment of the present invention.
Fig. 9 is a schematic upper surface view
showing a mask stage omitted in Fig. 8 and a
constitution of a mask stage measuring system.
Fig. 10 is a schematic upper surface view
showing a wafer stage omitted in Fig. 8 and a
constitution of a wafer stage measuring system.
Fig. 11 is a schematic view showing a
constitution of an exposure apparatus according to a
third embodiment of the present invention.
Fig. 12 is a schematic view showing a
constitution of an exposure apparatus according to a
fourth embodiment of the present invention.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
According to the present invention, an amount
of correction is obtained regarding at least one of a
mask and a photosensitive substrate, necessary for
substantial alignment between a pattern image formed
in the state of being moved from a reference image-forming
position and the photosensitive substrate,
based on the fluctuation amount of a reflecting
member disposed in a projection optical system from a
reference position. Then, by moving at least one of
the mask and the photosensitive substrate based on
the obtained amount of correction, the pattern image
and the photosensitive substrate are substantially
aligned with each other. In the described
specification, "reference image-forming position"
means the position of a pattern image formed by the
projection optical system in a state where the
reflecting member is located in the reference
position.
Specifically, an amount of positional shift and
an amount of rotational shift of the pattern image
from the reference image-forming position are
computed based on the fluctuation amount of the
reflecting member from the reference position, and a
correction amount of the mask is computed, which is
necessary for substantially correcting the amount of
positional shift and the amount of rotational shift.
Then, by moving only the mask based on the computed
amount of correction, the position shift and the
rotational shift of the pattern image are
substantially corrected, enabling substantial
alignment to be made between the pattern image and
the photosensitive substrate.
In addition, an amount of positional shift and
an amount of rotational shift of the pattern image
from the reference image-forming position are
computed based on the fluctuation amount of the
reflecting member from the reference position, and a
correction amount of the photosensitive substrate
necessary for substantially aligning the
photosensitive substrate with a pattern image is
computed, the pattern image having been formed in a
state where at least one of the amount of positional
shift and the amount of rotational shift occurs.
Then, moving only the photosensitive substrate based
on the computed amount of correction, the
photosensitive substrate is substantially aligned
with the pattern image formed in the state of being
moved from the reference image-forming position,
enabling substantial alignment to be made between the
pattern image and the photosensitive substrate.
Furthermore, an amount of positional shift and
an amount of rotational shift of the pattern image
are computed based on the fluctuation amount of the
reflecting member from the reference position. A
correction amount of the mask is computed, which is
necessary for substantially correcting one of the
amount of positional shift and the amount of
rotational shift. Also, a correction amount of the
photosensitive substrate is computed, which is
necessary for substantially aligning the
photosensitive substrate with a pattern image formed
in a state where the other of the amount of
positional shift and the amount of rotational shift
occurs. Then, by moving the mask based on the
computed correction amount of the mask to
substantially correct only the positional shift
(rotational shift) of the pattern image, and by
moving the photosensitive substrate based on the
computed correction amount of the photosensitive
substrate to substantially align the same with the
pattern image formed in the state of being shifted in
rotation (shifted in position) from the reference
image-forming position, substantial alignment can be
made between the pattern image and the photosensitive
substrate.
As can be understood from the foregoing,
according to the present invention, it is possible to
carry out alignment between the mask and a wafer even
during exposure as occasion demands, without
providing any surplus optical members in the optical
path between the mask and the wafer as the
photosensitive substrate, and according to the
positional or rotational shift of the pattern image
caused by the position fluctuation of the reflecting
member disposed in the projection optical system of a
catadioptric type. Therefore, even if fluctuation
occurs in the position of the reflecting member
during exposure to cause the positional or rotational
shift of the pattern image, by carrying out alignment
between the mask and the wafer during exposure as
occasion demands, it is possible to fabricate a good
semiconductor device always in the state of good
alignment.
Now, the preferred embodiments of the present
invention will be described with reference to the
accompanying drawings.
Fig. 1 schematically shows the constitution of
an exposure apparatus according to the first
embodiment of the present invention. In the first
embodiment, the present invention is applied to the
projection exposure apparatus of a scanning type,
which is designed to transfer a mask pattern to each
exposure area of the wafer by means of scanning
exposure while moving the mask and the wafer relative
to each other with respect to the projection optical
system of a catadioptric type.
In Fig. 1, a Z axis is set in parallel with the
optical axis AXa of the mask side portion 3a of a
projection optical system 3; a Y axis in parallel
with the paper surface of Fig. 1 within a surface
perpendicular to the Z axis; and an X axis
perpendicularly to the paper surface of Fig. 1 within
the surface perpendicular to the Z axis.
The exposure apparatus of Fig. 1 comprises an
illumination optical system 1 for uniformly
illuminating a mask 2 where a specified pattern
formed thereon to be transferred. In the
illumination optical system 1, for example, a light
emitted from an ArF excimer laser light source is
passed through a fly's eye lens to form a secondary
light source having a plurality of light source
images, and a light from this secondary light source
is converged through a condenser lens and emitted
toward the mask 2. Accordingly, by the light emitted
from the illumination optical system 1, the mask 2 is
uniformly illuminated in a superimposed manner. A
field of illumination (area of illumination) formed
on the mask 2 by the light emitted from the
illumination optical system 1 has a rectangular shape
extended thin and long in an X direction, and its
center is made eccentric in a - Y direction with
respect to the optical axis AXa of the mask side
portion 3a of the projection optical system 3.
The mask 2 is supported on a mask stage (not
shown) within an XY plane. The mask stage is
constituted of an XY stage two-dimensionally movable
within the XY plane, a rotary stage rotatable around
the Z axis, and so on. The mask stage is driven and
controlled by a driving system 12 based on a command
from a control system 11.
The light passed through the pattern of the
mask 2 is made incident on a concave mirror M3 along
a - Z direction through the mask side portion 3a of
the projection optical system 3. The light reflected
on the concave mirror M3 in a + Z direction is made
incident on a first reflecting mirror M1 disposed in
the optical path of the mask side portion 3a of the
projection optical system 3. The light reflected on
the first reflecting mirror M1 in a + Y direction is
made incident through two lens components on a second
reflecting mirror M2. The light reflected on the
second reflecting mirror M2 in the - Z direction is
transmitted through the wafer side portion 3b of the
projection optical system 3 to a wafer 4 as a
photosensitive substrate coated with a photosensitive
material such as a resist.
Accordingly, a pattern image of the mask 2 is
formed on the wafer 4. An exposure field formed on
the wafer 4 has a rectangular shape extended thin and
long in the X direction corresponding to the field of
illumination formed on the mask 2, and its center is
made eccentric in the + Y direction with respect to
the optical axis AXb of the wafer side portion 3b of
the projection optical system 3 (see Fig., 3,
described later). More specifically, the field of
illumination formed on the mask 2 and the exposure
field formed on the wafer 4 are similar to each other
in shape, and their sizes and eccentric distances
from the corresponding optical axes depend on the
projection magnification of the projection optical
system 3.
The wafer 4 is supported on a wafer stage
within the XY plane. As in the case of the mask
stage, the wafer stage is constituted of an XY stage
two-dimensionally movable within the XY plane, a
rotary stage rotatable around the Z axis, and so on.
Therefore, by carrying out scanning exposure
while moving the mask 2 and the wafer 4 in the Y
direction in synchronization with each other with
respect to the projection optical system 3, a mask
pattern can be transferred to one exposure area of
the wafer 4. In addition, by repeating the scanning
exposure while driving and controlling the wafer 4
two-dimensionally within the XY plane, mask patterns
can be transferred sequentially to the respective
exposure areas of the wafer 4 according to a so-called
step-and-scan system.
According to the first embodiment, the exposure
apparatus comprises a first detection system 13 for
detecting a fluctuation amount of the first
reflecting mirror M1 disposed in the projection
optical system 3 from the reference position, and a
second detection system 14 for detecting a
fluctuation amount of the second reflecting mirror M2
from the reference position. The outputs of the
first and second detection system 13 and 14 are
supplied to an arithmetic system 15. In the
arithmetic system 15, based on the fluctuation
amounts of the first and second reflecting mirrors M1
and M2 from the reference position respectively
detected by the first and second detection systems 13
and 14, a movement amount of a pattern image formed
through the projection optical system 3 from a
reference image-forming position, i.e., amounts of
positional and rotational shift are computed.
Subsequently, in the arithmetic system 15,
based on the computed amounts of positional and
rotational shift, a correction amount of the mask 2
is computed, which is necessary for correcting the
positional and rotational shift of the pattern image
from the reference image-forming position.
Information regarding the correction amount computed
in the arithmetic system 15 is supplied to the
control system 11.
In the control system 11, based on the
information of the correction amount supplied from
the arithmetic system 15, the mask stage is driven
and controlled by the driving system 12. In this way,
the mask 2 is moved by a required correction amount,
and the positional and rotational shift of the
pattern image from the reference image-forming
position is corrected. As a result, the pattern
image and the wafer 4 are aligned with each other.
Next, based on a specific constitution example,
the alignment between the mask 2 and the wafer 4 of
the first embodiment will be described.
Table (1) shows data values of the projection
optical system 3 of the first embodiment. In the
[entire data] of Table (1), NA represents a numerical
aperture of an image (wafer), and β represents a
projection magnification. In the [lens data] of
Table (1), the first column represents the order of
lens surfaces from the mask (object); r of the second
column represents a radius of curvature a lens
surface (∞ represents a plane); d of the third column
represents a spacing between lens surfaces; and n of
the fourth column represents a refractive index with
respect to the oscillation wavelength λ = 193.4 nm of
the ArF excimer laser.
First, as shown in Fig. 2, consideration is
given to a case where the first reflecting mirror M1
is shifted in rotation from the reference position by
1'' (1 sec.) clockwise around the Z axis. In this
case, as shown in Fig. 3, an exposure field EF to be
formed on the wafer 4 by being made eccentric in the
Y direction by + 7 mm with respect to the optical
axis AXb of the wafer portion 3b of the projection
system 3 is moved as a whole in the X direction by dX
= + 121 nm. The end portion image point P1 of the
exposure field EF away in the - X direction from the
center of the exposure field EF by 15 mm is moved in
the Y direction by dY = - 71 nm, and in the X
direction by dX = - 54 nm. In addition, the end
portion image point P2 of the exposure field EF away
in the + X direction from the center of the exposure
field EF by 15 mm is moved in the Y direction by dY =
+ 71 nm, and in the X direction by dX = - 54 nm.
Next, as shown in Fig. 4, consideration is
given to a case where the second reflecting mirror M2
is shifted in rotation from the reference position by
1'' clockwise around the Y axis. In this case, as
shown in Fig. 5, the exposure field EF formed on the
wafer 4 through the projection optical system 3 is
moved as a whole in the X direction by dX = - 794 nm.
The end portion image point P1 of the exposure field
EF away in the - X direction from the center of the
exposure field EF by 15 mm is moved in the Y
direction by dY = - 74 nm, and in the X direction by
dX = - 23 nm. In addition, the end portion image
point P2 of the exposure field EF away in the + X
direction from the center of the exposure field EF by
15 mm is moved in the Y direction by dY = + 74 nm,
and in the X direction by dX = - 23 nm.
Thus, when the first reflecting mirror M1 is
shifted in rotation by 1'' from the reference
position clockwise around the Z axis, and the second
reflecting mirror M2 is shifted in rotation by 1''
from the reference position clockwise around the Y
axis, the entire movement amount of the exposure
field EF and the movement amounts of the image points
P1 and P2 become as those shown in Table (2) as
synthesized movement amounts in Figs. 3 and 5.
| dY | dX |
Entire movement amount of EF | 0 nm | - 673 nm |
Movement amount of image point P1 | - 145 nm | - 77 nm |
Movement amount of image point P2 | + 145 nm | - 77 nm |
Specifically, when the first reflecting mirror
M1 is shifted in rotated by 1'' from the reference
position clockwise around the Z axis, and the second
reflecting mirror M2 is shifted in rotation by 1''
from the reference position around the Y axis, a
pattern image formed through the projection optical
system 3 is shifted in position by dX = - 673 nm in
the X direction, and shifted in rotation by 2''
clockwise around the Z axis.
As described above, according to the first
embodiment, by moving only the
mask 2 so as to
correct the positional and rotational shift of the
pattern image, alignment is performed between the
pattern image and the
wafer 4, and thus between the
mask 2 and the
wafer 4. Accordingly, in the specific
constitution example of the first embodiment, the
mask 2 is moved in the + X direction by 673 nm
(amount of positional shift of the pattern image) × 4
(inverse number of projection magnification) = 2692
nm, and rotated clockwise around the Z axis by an
angle equal to the rotational shift amount of the
pattern image, i.e., 2''. Thus, when only the
mask 2
is moved based on a required correction amount, the
entire movement amount of the exposure field EF and
the movement amounts of the image points P1 and P2
are corrected as shown in Table (3).
| dY | dX |
Entire movement amount of EF | 0 nm | 0 nm |
Movement amount of image point P1 | 0 nm | - 9 nm |
Movement amount of image point P2 | 0 nm | - 9 nm |
It can be understood from Table (3) that by
moving only the mask 2 based on a required correction
amount, the positional and rotational shift of the
pattern image caused by the rotational fluctuation of
the first and second reflecting mirrors M1 and M2
from the reference position can be corrected well.
Focusing on the movements of the image points P1 and
P2, the component dY of the Y direction is completely
corrected, while the component dX of the X direction
is left by a slight amount. However, optimization
can be performed to reduce the component dX of the X
direction by slightly deteriorating the component dY
of the Y direction.
As described above, in the specific
constitution example of the first embodiment, the
first detection system 13 detects the rotational
fluctuation angle of the first reflecting mirror M1
around the Z axis, and the second detection system 14
detects the rotational fluctuation angle of the
second reflecting mirror M2 around the Y axis. In
this case, for the detection system for detecting the
rotational fluctuation angle of the reflecting mirror,
for example, a constitution like that shown in Fig.
6A can be employed. In Fig. 6A, the detection system
like the first detection system 13 or the second
detection system 14 includes an illumination unit 62
for obliquely illuminating the end surface 61a of a
reflecting mirror 61 to be detected, and a detection
unit 63 for detecting the light receiving position of
a reflected light from the end surface 61a of the
reflecting mirror 61. The detection unit 63 detects
the rotational fluctuation angle of the reflecting
mirror 61 based on the light receiving position of
the reflected light from the end surface 61a of the
reflecting mirror 61. A detecting signal from the
detection unit 61 is supplied to the arithmetic
system 15.
The first and second detection systems 13 and
14 can employ constitutions respectively shown in
Figs. 6B and 6C. In the detection systems shown in
Figs. 6B and 6C, the illumination unit 62 obliquely
projects a light to the backside 61b of the
reflecting mirror 61, and the light receiving
position of a reflected light from the backside 61b
of the reflecting mirror 61 is detected by the
detection unit 63. As shown in Figs. 6B and 6C, the
rotational fluctuation and positional shift of the
reflecting mirror 61 are detected by two groups of
detection systems. In these variation examples as
shown in Fig. 6B, the detection system of the first
group includes, an illumination unit 62B having a
light source 621 and a lens 622 for collimating
lights from the light source 621, and a detection
unit 63B having a lens 631 for converging a light
reflected by the backside 61b of the reflecting
mirror and a photoelectric conversion device 632
disposed in the converging position of the lens 631.
The detection system of the second group
includes, an illumination unit 62C having a light
source 623 and a lens 624 for converging lights from
the light source 623 on the backside 61b of the
reflecting mirror, and a detection unit 63C having a
lens 633 for forming an image of the converging point
of the backside 61b of the reflecting mirror and a
photoelectric conversion device 634 disposed in the
image-forming position of the lens 633.
The rotational amount of the reflecting mirror
61 can be measured by monitoring the shift of the
converging position on the photoelectric conversion
device 632 of the detection unit 63B. The positional
shift of the backside 61b of the reflecting mirror in
a normal direction can be monitored by monitoring the
shift of the converging position on the photoelectric
conversion device 634 of the detection unit 63C.
On the surface of the reflecting mirror 61, a
reflection increasing film optimized for the
wavelength region of an exposure light is deposited.
On the backside 61b of the reflecting mirror 61, a
reflection increasing film optimized for the
wavelength region of a light (e.g., visible light)
generated by the illumination unit 62 should
preferably be deposited.
In the arithmetic system 15, the amounts of
positional and rotational shift of the pattern image
from the reference image-forming position are
computed based on the rotational fluctuation angles
of the first and second reflecting mirrors M1 and M2
respectively detected by the first and second
detection system 13 and 14. Then, in the arithmetic
system 15, based on the computed amounts of
positional and rotational shift of the pattern image,
the correction amount of the mask 2 is computed,
which is necessary for substantially correcting the
positional and rotational shift of the pattern image.
Information regarding the amount of correction
computed in the arithmetic system 15 is supplied to
the control system 11. The control system 11 drives
and controls the mask stage through the driving
system 12 based on the information of the correction
amount supplied from the arithmetic system 15. In
this way, the mask 2 is moved by a required
correction amount, and the positional and rotational
shift of the pattern image from the reference image-forming
position is substantially corrected. As a
result, good alignment can be made between the
pattern image and the wafer 4, and thus between the
mask 2 and the wafer 4.
After the alignment between the mask 2 and the
wafer 4 is completed in the foregoing manner, the
step moves to an exposure step (photolithography
step). Fig. 7 is a flowchart illustrating respective
steps for forming a specified circuit pattern on the
wafer by using the exposure apparatus of the present
invention.
First, in step 71 of Fig. 7, a metallic film is
vapor-deposited on the wafer of one lot. Then, in
step 72, a photoresist is coated on the metallic film
of the wafer. Thereafter, in step 73, by using the
exposure apparatus of the first embodiment, patterns
on the wafer are sequentially transferred through the
projection optical system to respective exposure
areas on the wafer. Then, in step 74, the
photoresist on the wafer is developed. Subsequently,
in step 75, by performing etching using the photo
resist pattern on the wafer as a mask, a circuit
pattern corresponding to the pattern on the wafer is
formed in each exposure area on each wafer. Then, a
circuit pattern of a layer further thereon is formed,
and the wafer passed through the wafer process is
sent to a next step.
After the end of the wafer process, in actual
assembling step, the steps including dicing for
cutting the wafer into chips for each baked circuit,
bonding for adding wiring to each chip, packaging for
each chip, and so on are performed, and a
semiconductor device such as an LSI or the like is
fabricated at the end.
The present invention has been described by
taking the example of fabricating the semiconductor
device such as an LSI by the photolithography process
in the wafer process using the exposure apparatus.
However, by the photolithography process using the
exposure apparatus, semiconductor devices, such a
flat panel display device (e.g., a liquid crystal
display device), a thin-film magnetic head, an
imaging device (CCD or the like), and so on can also
be fabricated.
As apparent from the foregoing, according to
the first embodiment, without providing any surplus
optical members in the optical path between the mask
2 and the wafer 4, alignment can be made between the
mask 2 and the wafer 4 even during exposure as
occasion demands, corresponding to the positional or
rotational shift of the mask pattern image caused by
the position fluctuation of the reflecting member M1
or M2 disposed in the projection optical system 3 of
the catadioptric type. As a result, even if
fluctuation occurs in the positions of the reflecting
members M1 and M2 during scanning exposure to cause
the mask pattern image to be shifted in position or
rotation, by carrying out alignment between the mask
2 and the wafer 4 during scanning exposure as
occasion demands, it is possible to fabricate a good
semiconductor device always in a well-aligned state.
According to the first embodiment, in the
arithmetic system 15, the amounts of positional and
rotational shift of the pattern image are computed
based on the rotational fluctuation angle of the
reflecting mirror, and the required correction amount
of the mask 2 is then computed based on the amounts
of positional and rotational shift of the pattern
image. However, without computing the movement
amount (amount of positional and rotational shift) of
the pattern image, the required correction amount of
the mask 2 can be computed directly from the
rotational fluctuation angle of the reflecting mirror.
In this case, the movement amount of the pattern
image and the required correction amount of the mask
2 can also be computed by, for example executing
computation based on a specified arithmetic
expression or referring to a specified memory table.
According to the first embodiment, the
alignment is carried out between the pattern image
and the wafer 4 by moving only the mask 2 to correct
the positional or rotational shift of the pattern
image. However, alignment can be made between the
pattern image and the wafer by moving only the wafer
4, and aligning the wafer 4 with the pattern image
formed by a movement from the reference image-forming
position. If only the wafer 4 is moved in the
foregoing specific constitution example, the wafer 4
is moved by 673 nm (amount of positional shift of the
pattern image) in the - X direction, and rotated by
an angle equal to the amount of rotational shift of
the pattern image, i.e., 2'', clockwise around the Z
axis.
Alternatively, alignment can also be made
between the pattern image and the wafer by moving
both of the mask 2 and the wafer 4. If the mask 2
and the wafer 4 are moved in the foregoing specific
constitution example, the mask 2 is moved by 2692 nm
in the + X direction to correct the positional shift
of the pattern image, and the wafer 2 is rotated by
2'' clockwise around the Z axis to be aligned with
the pattern image formed by rotational shift from the
reference image-forming position. Alternatively, the
mask 2 is rotated by 2'' counterclockwise around the
Z axis to correct the rotational shift of the pattern
image, and the wafer 4 is moved by 673 nm in the - X
direction to be aligned with the pattern image formed
by positional shift from the reference image-forming
position.
Furthermore, in the specific constitution
example of the first embodiment, the first detection
system 13 detects the rotational fluctuation angle of
the first reflecting mirror M1 around the Z axis, and
the second detection system 14 detects the rotational
fluctuation angle of the second reflection mirror M2
around the Y axis. Then, because of the rotational
fluctuation of the first reflecting mirror M1 around
the Z axis and the rotational fluctuation of the
second reflecting mirror M2 around the Y axis, the
pattern image is shifted in position in the X
direction and in rotation around the Z axis. However,
the fluctuation amount of the reflecting member from
the reference position detected by the detection
system is not limited to the rotational fluctuation
angle of the reflecting member around a specified
axis. The concept thereof includes the amount of
fluctuation, e.g., a rotational amount around the
other axis, a three-dimensional movement amount, and
so on.
Generally, if fluctuation occurs in the
rotation of one or a plurality of reflecting members
3 around three axes or fluctuation occurs three-dimensionally
in the position thereof, the pattern
image may be three-dimensionally shifted in position
or shifted in rotation around the three axes. In
such a case, the mask stage and the wafer stage must
be provided with a certain degree of freedom for
three-dimensional movement and a certain freedom for
rotation around the three axes. Specifically, in
addition to the XY stage and the rotary stage around
the Z axis, a Z stage movable in the Z direction and
a leveling stage rotatable around the X and Y axes
must be provided.
If the effect of the rotational fluctuation of
the reflecting member is particularly predominant,
then a rotational fluctuation angle around two axes
orthogonal to a normal thereof is detected on the
reflecting surface of the reflecting member. In this
case, a local coordinate is set, which is composed of
an X' axis parallel to the X axis and a Y' axis
parallel to the Y axis on the reflecting surface of
the first reflecting mirror M1 of the first
embodiment, and the rotational fluctuation angle of
the first reflecting mirror M1 around the X' axis and
the rotational fluctuation angle around the Y' axis
are detected. Similarly, a local coordinate is set,
which is composed of an X' axis parallel to the X
axis and a Y' axis parallel to the Y axis on the
reflecting surface of the second reflecting mirror M2
of the first embodiment, and the rotational
fluctuation angle of the second reflecting mirror M2
around the X' axis and the rotational fluctuation
angle around the Y' axis are detected.
Fig. 8 schematically shows a constitution of an
exposure apparatus according to a second embodiment
of the present invention. Fig. 9 is a schematic
upper surface view showing a mask stage omitted in
Fig. 8 and a constitution of a mask stage measuring
system. In addition, Fig. 10 is a schematic upper
surface view showing a wafer stage omitted in Fig. 8
and a constitution of a wafer stage measuring system.
The first and second embodiments are similar to
each other in constitution. The second embodiment is
different from the first embodiment only in that
alignment is carried out between the mask and the
wafer also by taking into consideration information
regarding the position of the mask stage and
information regarding the position of the wafer stage.
Accordingly, in Fig. 8, components having functions
like those of the components of the first embodiment
are denoted by reference numerals like those of Fig.
1. Now, the second embodiment will be described
focusing on the difference from the first embodiment.
In Fig. 9, the mask 2 is supported on a mask
stage 92 within the XY plane by a mask holder 91.
The mask stage 9 is consisted of an XY stage two-dimensionally
movable within the XY plane, a rotary
stage rotatable around the Z axis, and so on. The
rotational amount of the mask stage 92 on the XY
coordinate and around the Z axis is always measured
by a mask stage measuring system 16 (see Fig. 8).
In the mask stage measuring system 16, a mirror
93 having a reflecting surface extended in parallel
with the XZ plane and long in the X direction is
fixed on the mask stage 92. In addition, two corner
cube prisms 94 and 95 adapted to move beams made
incident in the + X direction in the Z direction in
parallel, and then emit the beams in the - X
direction are fixed on the mask stage 92 with a
spacing set therebetween in the Y direction.
Further, the mask stage measuring system 16
includes a laser interferometer 96 for interference
between a reflected beam obtained by making a beam
incident perpendicularly to the reflecting surface of
the mirror 93 and a reference beam. It also includes
laser interferometers 97 and 98 for interference
between reflected beams obtained by making beams
incident on the corner cube prisms 94 and 95 in the X
direction and reference beams. The laser
interferometers 96 to 98 are supported by proper
supports isolated from the mask stage 92. Thus, in
the mask stage measuring system 16, the Y direction
movement amount of the mask stage 92, i.e., a Y
coordinate, can be measured based on the output of
the laser interferometer 96. Also, based on the
outputs of the laser interferometers 97 and 98, the
XY direction movement amount of the mask stage 92,
i.e., an X coordinate, and the rotational amount of
the mask stage 92 around the Z axis can be measured.
Information regarding the position of the mask stage
92 measured in the mask stage measuring system 16 is
supplied to the arithmetic system 15.
In Fig. 10, the wafer 4 is supported on a wafer
stage 102 within the XY plane by a wafer holder 101.
The wafer stage 102 is consisted of an XY stage two-dimensionally
movable within the XY plane, a rotary
stage rotatable around the Z axis, and so on. The
rotational amount of the wafer stage 102 on the XY
coordinate and around the Z axis is always measured
by a wafer stage measuring system 17 (see Fig. 8).
In the wafer stage measuring system 17, a
mirror 103 having a reflecting surface extended in
parallel with the XZ plane and long in the X
direction is fixed on the wafer stage 102. In
addition, a mirror 104 having a reflecting surface
extended in parallel with the YZ plane and long in
the Y direction is fixed on the wafer stage 102.
Further, the wafer stage measuring system 17
includes a laser interferometer 105 for interference
between a reflected beam obtained by making a beam
incident perpendicularly to the reflecting surface of
the mirror 3 and a reference beam. The system also
includes a laser interferometer 106 for interference
between two reflected beams obtained by making
incident two beams having a spacing set therebetween
in the Y direction perpendicularly to the reflecting
surface of the mirror 104. Further, the system
includes a laser interferometer 107 for interference
between a reflected beam and a reference beam, the
reflected beam having been obtained by making a beam
incident toward the center of the exposure field EF
of the wafer 4 perpendicularly to the reflecting
surface of the mirror 104.
The laser interferometers 105 to 107 are
supported by proper supports isolated from the wafer
stage 102. Accordingly, in the wafer stage measuring
system 17, the Y direction movement amount of the
wafer stage 102, i.e., a Y coordinate, can be
measured based on the output of the laser
interferometer 105. Based on the output of the laser
interferometer 106, the rotational amount of the
wafer stage 102 around the Z axis can be measured.
Further, based on the output of the laser
interferometer 107, the X direction movement amount
of the wafer stage 102, i.e., an X coordinate, can be
measured. Information regarding the position of the
wafer stage 102 measured in the wafer stage measuring
system 17 is supplied to the arithmetic system 15.
According to the second embodiment, as shown in
Fig. 8, the outputs of the mask stage measuring
system 16 and the wafer stage measuring system 17 are
supplied together with the outputs of the first and
second detection systems 13 and 14 to the arithmetic
system 15. In the arithmetic system 15, the
correction amount of the mask 2 is computed, which is
necessary for substantial alignment between the mask
2 and the wafer 4 with respect to the projection
optical system 3. This computation is performed
based on the positional or rotational shift of the
pattern image from the reference image-forming
position computed from the rotational fluctuation
angles of the first and second reflecting mirrors M1
and M2 respectively detected by the first and second
detection systems 13 and 14, the information
regarding the position of the mask stage measured by
the mask stage measuring system 16 and the
information regarding the position of the wafer stage
measured by the wafer stage measuring system 17.
Information regarding the amount of correction
computed in the arithmetic system 15 is supplied to
the control system 11. The control system 11 drives
and controls the mask stage via the driving system 12
based on the information of the correction amount
supplied from the arithmetic system 15. Accordingly,
the mask 2 is moved by a required amount of
correction, the positional or rotational shift of the
pattern image from the reference image-forming
position is substantially corrected, and relative
positional shift between the mask 2 and the wafer 4
caused by relative positional shift between the mask
stage and the wafer stage is also corrected
substantially. As a result, the mask 2 and the wafer
4 are well aligned with each other with respect to
the projection optical system 3.
Therefore, as in the case of the first
embodiment, in the second embodiment, without
providing any surplus optical members in the optical
path between the mask 2 and the wafer 4, alignment
between the mask 2 and the wafer 4 can be performed
even during exposure as occasion demands, according
to the positional or rotational shift of the mask
pattern image caused by the positional fluctuation of
the reflecting member M1 or M2 disposed in the
projection optical system 3 of the catadioptric type,
and relative positional shift between the mask stage
and the wafer stage. As a result, even if
fluctuation occurs in the position of the reflecting
member M1 or M2 during scanning exposure to cause
positional or rotational shift of the mask pattern
image, or relative positional shift between the mask
stage and the wafer stage, by making alignment
between the mask 2 and the wafer 4 during scanning
exposure as occasion demands, a good semiconductor
device can be fabricated always in a well-aligned
state.
Similarly to the first embodiment, in the
second embodiment, by moving only the wafer 4 or both
of the mask 2 and the wafer 4, the mask 2 and the
wafer 4 can be aligned with each other with respect
to the projection optical system 3.
Each of the embodiments of the present
invention has been described by taking the example of
the exposure apparatus comprising the projection
optical system having two reflecting mirrors (optical
path deflecting member). However, the number and
disposition of reflecting mirrors should not be
limited to those of the first embodiment.
In each of the foregoing embodiments, the
optical system of the catadioptric type is applied as
a projection optical system. However, the present
invention can be applied to an optical system having
at least one reflecting member, for example, a
reflection type optical system, a refractive type
optical system having an optical path bending mirror,
and so on.
In addition, each of the embodiments of the
present invention has been described by taking the
example of the exposure apparatus of a so-called
step-and-scan system, which is designed to transfer a
mask pattern to each exposure area of the wafer by
executing scanning exposure while moving the mask and
the wafer in synchronization with each other with
respect to the projection optical system. However,
the present invention can also be applied to the
exposure apparatus of a so-called step and repeat
type, which is designed to sequentially transfer mask
patterns to the respective shot areas of the wafer
two-dimensionally while moving only the wafer by
stepping with respect to the projection optical
system, and the exposure apparatus of other types.
Furthermore, in each of the foregoing
embodiments, the ArF excimer laser light source is
used as a light source for the illumination optical
system. Instead, an F2 excimer laser light source
(wavelength 157 nm), a KrF excimer laser light source
(wavelength 248 nm), a mercury lamp for supplying an
i line (wavelength 365 nm) or one for supplying an
electromagnetic wave of soft X-ray wavelength region
of 13 nm can be used.
Fig. 11 is a schematic view showing a
constitution of an exposure apparatus according to a
third embodiment of the present invention.
The exposure apparatus shown in Fig. 11 is a
projection exposure apparatus for performing an
exposing operation by a step-and-scan system using
the light of a soft X - ray region (EUV light) having
a wavelength of 5 to 15 nm as an illumination light
for exposure. In Fig. 11, the optical axial
direction of a reflection reduction projection
optical system for forming a reduced image of a
reflection type mask 2 as a projection original on
the wafer 4 is set to be a Y direction; a direction
within a paper surface orthogonal to the Y direction
a Z direction; and a paper surface perpendicular
direction perpendicular to YZ direction an X
direction.
This projection exposure apparatus transfers
all the circuit patterns of the reflection type mask
2 respectively to a plurality of shot areas on the
wafer 4 by a step-and-scan system. Such transfer is
performed by one-dimensionally scanning the mask 2
and the wafer 4 relative to each other with respect
to a reflection reduction projection optical system 3
while projecting a part of the images of the circuit
patterns drawn on the reflection type mask 2 as
projection originals (reticle) through the reflection
reduction projection optical system 3 onto the wafer
4.
In this case, as the EUV light as an
illumination light for exposure in the described
embodiment has a low transmittance for atmosphere, an
optical path, through which the EUV light is passed,
is covered with a vacuum chamber, not shown, to be
shielded from external air.
First, an illumination optical system 1 shown
in Fig. 11 will be described.
A laser light source 100 has a function of
supplying a laser light having a wavelength of an
infrared region to a visible region, and for example,
a YAG laser or an excimer laser based on
semiconductor laser excitation can be used. This
laser light is converged by a converging optical
system 101 to be converged on a position 102. A
nozzle 103 injects a gaseous material to the position
102, and this injected material receives a high-illuminance
laser light at the position 102. At this
time, the injected material becomes high in
temperature because of the energy of the laser light,
excited in a plasma state, and emits an EUV light
during its transition to a low potential state.
Around the position 102, an elliptic mirror 104
is disposed to constitute the converging optical
system. This elliptic mirror 104 is positioned such
that its first focus can nearly coincide with the
position 102. On the inner surface of the elliptic
mirror 104, a multilayer film is formed to reflect an
EUV light. The EUV light reflected thereon is
converged at the second focus of the elliptic mirror
104, and then passed toward a parabolic mirror 105
constituting the converging optical system. The
parabolic mirror 105 is positioned such that its
focus can nearly coincide with the second focus
position of the elliptic mirror 104. On the inner
surface thereof, a multilayer film is formed to
reflect an EUV light.
An EUV light emitted from the parabolic mirror
105 is passed in a nearly collimated state toward
reflection type fly's eye optical systems 106a and
106b as optical integrators. The reflection type
fly's eye optical systems 106a and 106b are composed
of a first reflection element group 106a integrating
a plurality of reflecting surfaces, and a second
reflection element group 106b having a plurality of
reflecting surfaces corresponding to the plurality of
reflecting surfaces of the first reflection element
group 106a. On the plurality of reflecting surfaces
constituting each of these first and second
reflection element groups 106a and 106b, multilayer
films for reflecting EUV lights are also provided.
The collimated EUV light from the parabolic
mirror 105 is divided for its wave front by the first
reflection element group 106a, and EUV lights from
the respective reflecting surfaces are converged to
form a plurality of light source images. The
plurality of reflecting surfaces of the second
reflection element group 106b is positioned
respectively near the forming positions of the
plurality of light source images. The plurality of
reflecting surfaces of the second reflection element
group 106b substantially function as field mirrors.
Accordingly, the reflection type fly's eye optical
systems 106a and 106b form a number of light source
images as secondary light sources based on
approximate parallel beams from the parabolic mirror
105. Such a reflection type fly's eye optical system
106 is disclosed in Japanese Patent Application No.
10-47400 filed by the applicant of the present
invention. In the described embodiment, to control
the shape of the secondary light source, a σ
diaphragm should preferably be provided near the
second reflection element group 106b.
EUV lights emitted from the secondary light
sources formed by the reflection type fly's eye
optical systems 106a and 106b are passed toward a
condenser mirror 107 positioned so as to set a focus
position near the positions of the secondary light
sources. After being reflected and converged by the
condenser mirror 107, the EUV lights are passed
through an optical path bending mirror 108 to reach
the reflection type mask 2. On the surfaces of the
condenser mirror 107 and the optical path bending
mirror 108, multilayer films are formed to reflect
the EUV lights. The condenser mirror 107 converges
the EUV lights emitted from the secondary light
sources, and uniformly illuminates the reflection
type mask 2 in a superimposed manner.
In the described embodiment, to spatially
separate an optical path between an illumination
light directing the reflection mask 2 and an EUV
light reflected by the reflection type mask 2 to
direct the projection system, the illumination system
is set as a non-telecentric system, and the
reflection reduction projection optical system 3 is
also set as a mask side non-telecentric optical
system.
On the reflection type mask 2, a reflection
film composed of a multilayer film for reflecting an
EUV light is provided. This reflection film has a
pattern corresponding to the shape of a pattern to be
transferred onto the wafer 4 as a photosensitive
substrate. An EUV light reflected by the reflection
mask 2 and provided with the pattern information of
the reflection mask 2 is made incident on the
reflection reduction projection optical system 3.
The reflection reduction projection optical
system 3 is constituted of a mirror M1 as an optical
path bending mirror, a concave mirror M2, a convex
mirror M3, and a concave mirror M4. The mirrors M1
to M4 constituting the reflection reduction
projection optical system 3 have multilayers provided
on the substrate to reflecting EUV lights.
The EUV light reflected by the reflection type
mask 2 is passed through the reflection reduction
projection optical system 3 to form, in the exposure
area of circular arc shape on the wafer 2, a reduced
image of the pattern of the reflection type mask 2 by
a specified reduction magnification β (e.g., |β| = 1/4,
1/5, 1/6).
The reflection type mask 2 is supported by a
mask stage 2S movable at least in the Y direction and
in a rotational direction around the Z axis. The
wafer 4 is supported by a wafer stage 4S movable in
the XYZ direction. The movements of these mask and
wafer stages 2S and 4S are achieved respectively by a
mask stage driving system 12 and a wafer stage
driving system (not shown). During an exposing
operation, while illuminating the reflection type
mask 2 with an EUV light by the illumination system 1,
the reflection type mask 2 and the wafer 4 are moved
with respect to the reflection type projection
optical system 3 at a specified speed ratio set by
the reduction magnification of the projection system
3. Thus, the pattern of the reflection type mask 2
is scanned and exposed within a specified shot area
on the wafer 4.
According to the embodiment, the rotational
amount of the mask stage 2S on the XY coordinate and
around the Z axis is constantly measured by the mask
stage measuring system 16, and information regarding
the measured position of the mask stage 2S is
supplied to the arithmetic system 15. In addition,
the rotational amount of the wafer stage 4S on the XY
coordinate and around the Z axis is constantly
measured by the wafer stage measuring system 17, and
information regarding the measured position of the
wafer stage 4S is supplied to the arithmetic system
15.
The rotational fluctuation angle and the
positional shift amount of the mirror M1 of the
reflection reduction projection optical system 3 as a
projection system are detected by the detection
system 13, and outputted to the arithmetic system 15.
In the arithmetic system 15, similarly to the
foregoing second embodiment, the correction amount of
the mask 2 is computed, which is necessary for
substantial alignment between the mask 2 and the
wafer 4 with respect to the projection system 3.
This computation is performed based on the positional
and rotational shift of the pattern image from the
reference image-forming position obtained based on
the rotational fluctuation angle and the amount of
positional shift of the reflecting mirror M1 detected
by the detection system 13, the position information
of the mask stage 4S measured in the mask stage
measuring system 16, and the position information of
the wafer stage measured in the wafer stage measuring
system.
Information regarding the amount of correction
computed in the arithmetic system 15 is supplied to
the control system 11. The control system 11 drives
and controls the mask stage via the driving system 12
based on the information of the correction amount
supplied from the arithmetic system 15. In this way,
the mask 2 is moved by a required amount of
correction to substantially correct the positional or
rotational shift of the pattern image from the
reference image-forming position. Also, relative
positional shift between the mask 2 and the wafer 4
caused by the relative positional shift between the
mask stage and the wafer stage is substantially
corrected. As a result, the mask 2 and the wafer 4
are well aligned with each other with respect to the
projection system 3.
As in the cases of the first and second
embodiments, in the third embodiment, by moving only
the wafer 4 or both of the mask 2 and the wafer 4,
alignment can be made between the mask 2 and the
wafer 4 with respect to the projection optical system
3.
In each of the foregoing embodiments, the
rotational fluctuation angle of the plane mirror as
an optical path deflecting member (optical path
bending mirror) in the projection optical system was
detected. However, the present invention can be
applied even if a mirror for detecting the rotational
fluctuation angle has power (e.g., a concave or
convex mirror), as long as the mirror bends the
optical axis. In such a case, not only the
positional fluctuation or rotational shift of the
image but also the change of an aberration state may
occur. However, if the change of the aberration
stage is within a permissible amount, by moving the
position of at least one of the mask and the
photosensitive substrate in this state, exposure can
be realized under a good aligned state and an
aberration corrected state. In addition, if the
change of the aberration state exceeds the
permissible amount, then the aberration state is
corrected by moving a part of the optical member
constituting the projection optical system.
Fig. 12 is a schematic view showing a
constitution of an exposure apparatus according to
the fourth embodiment of the present invention.
The exposure apparatus shown in Fig. 12 is a
projection exposure apparatus designed to carry out
exposure by projecting the unmagnified (i.e. the same
size) image of a circuit pattern formed on the mask 2
onto the photosensitive substrate 4 coated with a
photosensitive material on a glass substrate, and
moving the mask 2 and the photosensitive substrate 4
relative to each other with respect to an
unmagnifying projection optical system 3. The
projection exposure apparatus based on such a system
is disclosed, for example in US Patent No. 5298939.
In Fig. 12, a coordinate system is set such
that the circuit pattern forming surface of the mask
2 can becomes an XY plane. An X axis is set in a
direction perpendicular to a paper surface; a Y axis
in a direction within the paper surface; and a Z axis
in a direction orthogonal to the XY plane and within
four surfaces.
The exposure apparatus shown comprises an
illumination optical system for uniformly
illuminating the mask 2 having patterns formed
thereon to be transferred. As one similar to a
system described in US Patent No. 5298939 can be used
for this illumination optical system, explanation
thereof will be omitted. A field of illumination
(illumination area) formed on the mask 2 by the
illumination optical system 1 can be formed in a
trapezoidal shape having, for example a base extended
in the X direction and an upper side in a - Y
direction side from the base.
The mask 2 is supported, as in the case of the
foregoing embodiments, by a mask stage 2S having an
XY stage two-dimensionally movable within the XY
plane and a stage (rotary stage) rotatable around
the Z axis.
The projection optical system 3 consists of a
first image-forming system 3A for forming an
unmagnified intermediate image of the mask 2, and a
second image-forming system 3B for forming the
intermediate image again on the photosensitive
substrate 4 in unmagnified state. The first image-forming
system 3A consists of an optical path bending
mirror M1, a lens group 3A1 and a concave mirror M3.
The second image-forming system 3B consists of an
optical path bending mirror M2, a lens group 3B1 and
a concave mirror M4. A light transmitted through the
pattern of the mask 2 is reflected by the optical
path bending mirror M1, and proceeds in the + Y
direction. Then, the light is passed through the
lens group 3A1 and reflected by the concave mirror M3.
And then, the light is passed again through the lens
group 3A1 and reflected by the optical path bending
mirror M1, proceeds in the - Z direction, and then
forms an unmagnified intermediate image (lateral
magnification = - 1) of the mask 2. A light from
this intermediate image is reflected by the optical
path bending mirror M2, and proceeds in the + Y
direction. Then, the light is passed through the
lens group 3B1 and reflected by the concave mirror M4.
The light is then passed through the lens group 3B1
again, reflected by the optical path bending mirror
M2, and proceeds in the - Z direction to form an
unmagnified secondary image (lateral magnification =
+ 1) of the mask 2 on the photosensitive substrate 4.
The photosensitive substrate 4 is supported, as
in the case of the foregoing embodiments, by a
substrate stage 4S having an XY stage two-dimensionally
movable within the XY plane and a
stage (rotary stage) rotatable around the Z axis.
This substrate stage 4S is driven and controlled by
the driving system 12 based on a command from the
control system 11.
Thus, by carrying out scanning exposure while
moving the mask 2 and the photosensitive substrate 4
integrally in the Y direction with respect to the
projection optical system 3, the mask pattern can be
transferred to an exposure area extended in the Y
direction on the photosensitive substrate 4.
Subsequently, by moving the mask 2 and the
photosensitive substrate 4 integrally in the X
direction (stepping operation) and repeating the
foregoing scanning exposure, the mask patterns can be
transferred to all the areas on the photosensitive
substrate 4.
According to the fourth embodiment, the
exposure apparatus comprises a first detection system
13 for detecting the fluctuation amount of the first
reflecting mirror M1 disposed as an optical path
bending mirror disposed in the projection optical
system 3 from the reference position, and a second
detection system 14 for detecting the fluctuation
amount of the second mirror M2 as an optical path
bending mirror from the reference position.
Similarly to the second embodiment, in the fourth
embodiment, the exposure apparatus further comprises
a mask stage measuring system 16 for measuring the
coordinate position of the mask stage 2S, and a
substrate stage measuring system 17 for measuring the
coordinate position of the substrate stage 4S. The
outputs of the first and second detection systems 13
and 14 and the mask and substrate stage measuring
systems 16 and 17 are supplied to the arithmetic
system 15. As in the case of the second embodiment,
the arithmetic system 15 computes the correction
amount of the photosensitive substrate 4, which is
necessary for substantial alignment between the mask
2 and the photosensitive substrate 4 with respect to
the projection optical system 3. This computation is
performed based on the amount of positional or
rotational shift of the pattern image from the
reference image-forming position obtained based on
the rotational fluctuation angles of the first and
second reflecting mirrors M1 and M2 measured by the
first and second measuring systems 13 and 14,
information regarding the position of the mask stage
2S measured by the mask stage measuring system 16,
and information regarding the position of the
substrate stage measured by the substrate measuring
system 17.
Information regarding the amount of correction
computed in the arithmetic system is supplied to the
control system 11. The control system 11 drives and
controls the substrate stage 4S via the driving
system 12 based on the information of the correction
amount supplied from the arithmetic system 15. In
this way, the photosensitive substrate 4 is moved by
a required amount of correction, and the positional
or rotational shift of the pattern image from the
reference image-forming position is substantially
corrected. Also, relative positional shift between
the mask 2 and the photosensitive substrate 4 caused
by relative positional shift between the mask stage
2S and the substrate stage 4S is substantially
corrected. As a result, the mask 2 and the
photosensitive substrate 4 are well aligned with each
other with respect to the projection optical system 3.
According to the fourth embodiment, a large-area
exposure area is obtained by joining a plurality
of unit exposure areas for exposure. In this case,
to further enhance the accuracy of joining among the
unit exposure areas, during scanning exposure, the
moving direction of at least one of the mask 2 and
the photosensitive substrate 4 should not be limited
in the Y direction, but the moving amount component
of the + X or - X direction should preferably be
added according to the rotational direction of the
pattern image.
In each of the foregoing embodiments, the
position correction of the mask and the
photosensitive substrate is constantly performed
according to the positional or rotational shift of
the pattern image caused by the positional
fluctuation of the reflecting member disposed in the
projection optical system. However, such correction
needs not be carried out constantly.
For example, correction may be carried out when
movement to a next shot area after exposure for one
shot area on the photosensitive substrate, at the
time of replacement of the photosensitive substrate,
or at the time of mask replacement.
Also, the position fluctuation of the
reflecting member is constantly monitored by the
detection system, and correction may be carried out
when the amount of positional fluctuation of the
reflecting member exceeds a specified permissible
amount. This variation example will now be described
by referring to, for example Fig. 11. In Fig. 11, a
permissible amount for the positional fluctuation of
the reflecting member is stored beforehand in the
arithmetic system 15. Determination is made by the
arithmetic system 15 as to whether the amount of
positional fluctuation detected by the detection
system 14 has reached this permissible amount or not.
If the amount of positional fluctuation is within the
permissible amount, then no position correction is
performed for at least one of the mask and the
photosensitive substrate. If the amount exceeds the
permissible amount, an amount of correction is
computed for at least one of the mask and the
photosensitive substrate based on the amount of
positional fluctuation, and the selected one of the
mask and the photosensitive substrate is moved. If
the amount of positional fluctuation exceeds the
permissible amount, a display 18 as a display unit
connected to the arithmetic system 15 should
preferably display that the permissible amount has
been exceeded.
POSSIBLE INDUSTRIAL APPLICATION
As apparent from the foregoing, according to
the exposure apparatus and the exposure method of the
present invention, without providing any surplus
optical members in the optical path between the mask
and the wafer as a photosensitive substrate, it is
possible to carry out alignment between the mask and
the wafer even during exposure as occasion demands
according to the positional or rotational shift of
the pattern image caused by the positional
fluctuation of the reflecting member disposed in the
projection optical system of the catadioptric type.
Therefore, even if the pattern image is shifted in
position or rotation because of fluctuation in the
position of the reflecting member during exposure, by
carrying out alignment between the mask and the wafer
during exposure as occasion demands, it is possible
to fabricate a good semiconductor device always in a
well-aligned state.